Accurate forecasting of bio-syngas composition via ANN in catalytic biomass gasification

Abstract Catalytic biomass gasification is a process that involves the conversion of solid biomass constituents into a gaseous fuel, typically a mixture of CO, H2, CO2, and CH4 through high-temperature thermochemical reactions in the presence of a catalyst. In this research, the MIMO (multiple-input multiple-output) layer Artificial Neural Network (ANN) models have been implemented to predict the individual gas yields in the catalytic biomass gasification process. The experimental data consisting of 315 fuel samples is collected from coal, biomass and coal-biomass blends. The dataset is comprised of the compositional analysis (proximate and ultimate analysis) and from different types of gasifiers (updraft, downdraft, fluidised bed and entrained bed). Catalytic biomass gasification done under conditions (Temperature range 700–900 °C, equivalence ratio 0.3, and Ash as a catalyst). ANN models having 11 input features (C, H, N, S, O, MC, Ash, T, VM, LHV and ER) and 5 output features (CO, CO2, CH4 & H2 and gas yield) are designed with multiple neurons in the hidden layers, a denser to predict catalytic gasification of biomass. The ANN models are trained by Levenberg-Marquardt (LM) back-propagation and Bayesian regularization (BR) algorithms. The individual gas yields predicted by the ANN models are compared by using mean square error and Regression analysis. It has been observed that the LM algorithm produced better results than BR. The ANN model provided results are in good agreement with experimental data.